Changelog
v1.7.3 (2025-03-17)¶
Fixes¶
- Fixed a wrong configuration where assets were not considered in the Kubernetes deployment.
- Fixed a wrong configuration in the Docker deployment where the OIDC provider did not consider other URLs than
http://localhost
.
v1.7.2 (2025-03-13)¶
Fixes¶
- Fixed a wrong configuration of
caffeine
in the Data Service that did not find views/subsets after table creation within the cache period of 60 seconds in #506.
v1.7.1 (2025-03-06)¶
Features¶
- Added support to download
pandas
DataFrame by PID in #503. - Added the possibility to create and fill a table from a
pandas
DataFrame (or optionally just create the schema) in #496.
Fixes¶
- Fixed a bug where quick interaction with the UI caused the user to trigger the brute-force login detection in #501.
v1.7.0 (2025-03-03)¶
Contains Breaking Changes
This release updates the Metadata Database schema which is incompatible to v1.6.3! Follow the steps:
- Make a backup of the database with
mariadb-dump
. - Apply the schema changes script:
schema.sql
:mariadb -h 127.0.0.1 -p3306 -u root --password=<password> -D dbrepo < schema.sql
- Install the dependencies from the
requirements.txt
file or use your local environment:pip install dbrepo==1.6.5rc15
- Run the data migration script
data.py
:It generates the SQL statements used for migrating to the new schema.python data.py > data.sql
- Run the generated
data.sql
script:mariadb -h 127.0.0.1 -p3306 -u root --password=<password> -D dbrepo < data.sql
Features¶
- Implemented a basic brute-force security defense strategy in the Auth Service that increments the wait time on wrong logins in #494.
- Implemented a password policy in #495.
Changes¶
- Replaced sequential numerical ids with non-guessable random ids in the Metadata Database in #491.
- Changed the interface for executing query in subsets/views in #493.
Removals¶
- Removed the Upload Service in favor of an internal stable upload endpoint in the Data Service in #492.
v1.6.5 (2025-02-18)¶
Fixes¶
- Fixed a bug where listing the views in the Python library did not work.
- Fixed a wrong MariaDB configuration where the
innodb_buffer_pool_size
variable was not configured to 70% of the available memory in the Helm chart.
v1.6.4 (2025-02-14)¶
Fixes¶
- Fixed a bug where the users were not synced with the Metadata Database in #489.
v1.6.3 (2025-02-05)¶
Changes¶
- Refactored the UI to support OIDC and added an event listener to the Auth Service that syncs users on creation to the Metadata DB in #488.
v1.6.2 (2025-01-24)¶
Changes¶
- Added interface tests for the Python library in Gitlab CI/CD pipeline in #486.
Fixes¶
- Fixed a bug where no pagination was possible in #487.
v1.6.1 (2025-01-21)¶
Changes¶
- Added privacy feature for hidden databases (and optionally tables, views, subsets) that hides them completely from e.g. the search in #482.
Fixes¶
- Added init container that adds the admin user to the Metadata Database in #480.
v1.6.0 (2025-01-07)¶
Features¶
- Added possibility to modify table description and privacy mode that hides metadata of databases, tables, subsets and views in #472.
- Added Auth Service init container to sync internal system ("admin") user into Metadata Database in #470
Changes¶
- Drop MariaDB Galera chart version from
14.0.12
to13.2.7
because of stability issues with the newer versions on the OpenShift Kubernetes cluster. - Bumped Grafana chart version from
10.1.1
to11.4.2
. - Bumped NGINX chart version from
18.2.6
to18.3.1
. - Bumped SeaweedFS chart version from
1.0.2
to4.2.1
(created a pull request #31192). - Improve RESTfulness of HTTP API to e.g. not include the database image due to size in #463.
Fixes¶
- Fixed a wrong JPA entity configuration in Hibernate that in some cases causes the
SERIAL
increment to not trigger in MariaDB Galera. Removed the deprecated@GenericGenerator
annotation on JPA entities. - Fixed a bug where the dataset separator was being ignored for imports in #478.
v1.5.3 (2024-12-13)¶
Fixes¶
- Fixed a bug where subsets containing sub-queries are not able to retrieve data in #476.
v1.5.2 (2024-12-03)¶
Changes¶
- Adapt Helm chart to support
runAsNonRoot
throughout and specifyresource
presets for the highly-constrained OpenShift Kubernetes environment in #467. - Require authentication for uploading files in #466.
Fixes¶
- Fixed a validation problem failing to validate UUIDs in #471.
- Fixed the
dist.tar.gz
file not being found in the CI/CD pipeline onrelease-
branches in #465.
v1.5.1 (2024-11-09)¶
Fixes¶
- Bug where the data volume could not be calculated when the data length column in the Metadata Database is
null
in #462. - Bug where the schema could not be created manually in #461.
v1.5.0 (2024-11-06)¶
Contains Breaking Changes
This release updates the Metadata Database schema which is incompatible to v1.4.6! Use the migration
script schema_1.4.5-to-1.5.0.sql
to apply the changes manually.
Features¶
- Added
SERIAL
data type to create incrementing key in #454
Changes¶
- Remove the Data Database Sidecar and replace it with Apache Spark in #458.
- Allow anonymous users to create subsets for public databases in #449.
- Show file upload progress in #448.
- Change the Docker image of the Auth Service to Bitnami-maintained similar to Kubernetes deployment with accompanying Auth Database change to PostgreSQL in #455
Fixes¶
- Preventing the semicolon
;
to be used in UI and fixed cryptic subset execution error messages in #456. - Multiple UI errors in #453.
- Fixed install script.sh in #444
- No hardcoded data type metadata in UI but instead added it hardcoded (associated with
image_id
) Metadata Database.
v1.4.6 (2024-10-11)¶
Contains Breaking Changes
This release updates the Metadata Database schema which is incompatible to v1.4.5!
Features¶
- Added Dashboard Service and monitoring in default setup.
Changes¶
- Show the progress of dataset uploads in the UI in #448
- Anonymous users are allowed to create (non-persistent) subsets in #449
- Removed logic that maps
True
,False
andnull